Reproduce “Coarse-to-Fine Learning for Multi-Pipette Localisation in Robot-Assisted In Vivo Patch-Clamp” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lan Wei, Gema Vera Gonzalez, Phatsimo Kgwarae, Alexander Timms, Denis Zahorovsky, Simon Schultz, Dandan Zhang
The method localizes multiple pipette tips in noisy two-photon images for robot-assisted patch-clamp. It enhances image crops with CycleGAN, predicts coarse regions with a Vision Transformer, and refines coordinates with a ResNet-18 decoder. On 341 in-vivo images, it achieves 98.62% point accuracy within 10 μm, versus ViT's 84.13%; images show pipettes before brain entry, not in-tissue navigation or recording.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.